AI Command Center

A dependency-free LLM gateway and self-hosted usage & cost dashboard. You point any project at it - any language, one command - and it captures tokens, cost, latency, and errors for every AI product in one place.

The idea in one picture

Your apps already call an LLM through a provider SDK. Every SDK lets you set a custom base URL. Point that base URL at the gateway and it forwards the request to the real provider untouched (your API key included), streams the response straight back, and reads the token usage on the side to compute cost. Nothing else about your code changes.

your app  ──base_url──▶  AI Command Center gateway  ──▶  OpenAI / Anthropic / Gemini / …
                              │
                              ├─ logs tokens, cost, latency, errors  (metadata only)
                              └─ serves the dashboard at :4321

What it is - and isn't

It is: a self-hosted, language-agnostic cost/usage dashboard that runs from a single command with no database and no external services. Your provider keys pass through; prompt and response bodies are never stored.

It isn't: a tracing platform, an eval framework, a prompt manager, or a routing/failover gateway. If you need those, the comparison points you to the right tool.

Next steps

• Install & run - get the gateway up in under a minute.
• Integrate a project - snippets for every language.
• Auth & teams - lock it down for shared use.
• HTTP API - build on the raw JSON endpoints.